Debug and Testing
Use the Output Log category InventorySubsystem and inspect the Pawn, component identity tag, item lookup and slot GUID involved in a failing request. RunOperation failures include the operation tag, FailReason, RequestedAmount and AppliedAmount.
Functional Checks
| Case | Verify |
|---|---|
| Fresh startup | Configured item/database lookups become valid; missing assets show an error |
| Add 1 / add past stack limit | Positive stack counts, new GUID per new entry, correct OutAddedAmount |
| Full slot inventory | Existing stack space remains usable; additional new slots are restricted |
| Weight and Hybrid limits | Full and partial adds respect the configured maximum |
| Remove available amount | Quantity decreases and empty entries disappear |
| Invalid Pawn/ID/tag/GUID/amount | Failure is visible and the integration handles it |
| Partial pickup | Remaining world quantity is preserved |
| Open/close UI and late binding | Snapshot includes existing entries; no duplicate widgets |
| Respawn/travel | New Pawn's components are resolved and UI is rebound |
Network Checks
Use at least two clients with listen-server and dedicated-server PIE modes. Verify client RPC ownership, server validation, late join, actor relevancy and the local listen-server UI. Repeat under latency/loss simulation in the editor and in a packaged build.
Exercise the cases in Implementation Limits, especially over-removal, specific removal, swap, external transfer, slot-add carry weight and rejected visual prediction. A Failed result is not a universal rollback guarantee.
The source exposes logging and example component/widgets. The reviewed content does not establish a complete automated test suite or native debug inventory actor. The documentation check itself did not run Unreal tests.